Di He

Di He

Senior Data Scientist

Fidelity Investments

Biography

Di is a senior data scientist at Fidelity Investments. She have 3-years data related working experiences in Finance, Consulting and Retail industry. Her projects includes Image Classification, Recommendation System, Question & Answering and Reinforcement Learning in Portfolio Management. She is passionate in applying data science in real world and make her own impact.

Download my resumé.

Interests
  • Machine Learning
  • Deep Learning
  • NLP
  • Reinforcement Learning
Education
  • MSc in Data Science, 2020 - 2021

    New York University

  • BA in Accounting, 2013 - 2017

    Xiamen University

Experience

 
 
 
 
 
Senior Data Scientist, Advanced Technology in Investment Management
Jan 2022 – Present Jersey City, NJ
 
 
 
 
 
Data Science Summer Intern, Advanced Technology in Investment Management
Jun 2021 – Aug 2021 Boston, MA
  • Measured equity’s risk exposures on market risk factors by correlation and mutual information. Researched on change point detection models to identify equity return pattern changes based on time series of its risk exposures. On the top of relevant periods, trained XGBoost models to predict equity returns. Automated the whole pipeline on AWS via Spark for 10000+ assets using 30 years’ daily data.
  • Imputed quarterly data to daily data using gradient-based optimization on top of PyTorch.
 
 
 
 
 
Data analyst, Retail Data Science
Jul 2019 – Jun 2020 Shanghai, China
  • Predicted product-level sales for new season using random forests and interpreted it with SHAP. Built a website using Dash and deployed on Heroku to help users interact with model outcomes.
  • Recommended new store locations for expansion team using k-means and DBSCAN Clustering with parsed transportation and competitor stores’ geolocation data getting from Gaode map API.
  • Designed and maintained tables in Hive and built Tableau dashboards for stores and products.
  • Forecasted monthly sales via exponential smoothing and Facebook prophet to help set daily sales targets. Allocated them daily based on holiday and week seasonality measured by regression.
 
 
 
 
 
Business Analyst, Enterprise Application
Jul 2017 – May 2019 Shanghai, China
  • As a consultant, prepared requirements documents and function specifications documents for client’s financial information system. Designed prototypes using Axure RP and tested final applications.
  • Increased client’s operation efficiency by building management dashboards for 200+ subsidiaries.

Skills

Python (pandas, matplotlib, scikit-learn)

90%

Torch

70%

Spark

60%

AWS EMR

50%

SQL

90%

Tableau

80%